Survey of Viable Airborne Fungi in Wine Cellars of Tokaj, Hungary

Aerobiologia https://doi.org/10.1007/s10453-017-9505-3

ORIGINAL PAPER

Survey of viable airborne fungi in wine cellars of Tokaj, Hungary

Dona´t Magyar . Zolta´nKa´llai . Matthias Sipiczki . Csaba Dobolyi . Flo´ra Sebok} . Tı´mea Beregsza´szi . Zolta´n Bihari . La´szlo´ Kredics . Gyula Oros

Received: 25 May 2017 / Accepted: 29 November 2017 Ó Springer Science+Business Media B.V., part of Springer Nature 2017

Abstract The composition of fungal biota and air compositions; however, their dominant species were quality of five traditional subterranean wine cellars proved to be the same. Among the isolated strains and one store building of a modern wine production Penicillium spp. were the most frequent. The walls of facility were examined in the Tokaj wine region cellars were covered by colonies of Zasmidium (northeastern Hungary). Air samples were collected (Cladosporium) cellare often referred to as a noble with SAS IAQ sampler onto PDA, MEA and RBA. mold. Even so, this mold has been found only at a Strains representing morphotypes were isolated from small concentration in the air samples (10–30 CFU/ colonies formed on agar plates from either air or m3). The walls of the modern store were free of molds. surface samples. The internal transcribed spacer (ITS) Diversity of fungi of the examined wine cellars was region of the rRNA gene cluster was amplified with influenced by environmental conditions to a certain primers ITS1 and ITS4. Altogether 90 morphotypes degree, such as elevation (height above sea level), age, were isolated, 48 and 12 strains (43 species) from the reconstruction time of cellars, indoor ethanol concen- air and surfaces, respectively. The number of spore- tration and the number of chimneys. The location of forming species generated high diversity of indoor cellars poorly influenced the concentration of fungi of fungi and differences between the cellars’ fungal the air inside cellars, contrary to outdoors where the air

D. Magyar (&) T. Beregsza´szi Department of Air Hygiene and Aerobiology, National Ministry of Human Capacities, Budapest, Hungary Public Health Institute, Albert Flo´ria´n Str. 2–6, Budapest, Hungary L. Kredics e-mail: [email protected] Department of Microbiology, Faculty of Science and Informatics, University of Szeged, Szeged, Hungary Z. Ka´llai Á Z. Bihari Research Institute for Viticulture and Oenology, Tokaj, G. Oros Hungary Plant Protection Institute, Hungarian Academy of Sciences, Budapest H-1022, Hungary Z. Ka´llai Á M. Sipiczki Department of Genetics and Applied Microbiology, University of Debrecen, Debrecen, Hungary

C. Dobolyi Á F. Sebok} Institute of Aquaculture and Environmental Safety, Szent Istvań University, Go¨do¨llo,} Hungary 123 Aerobiologia of the municipal area contained more CFUs than that temperatures of 10–12 °C (except for the differences of rural spaces. in summer and winter temperatures of ca. 1.5 °C) as well as air humidity of 85–95% (Gasˇinec et al. 2012). Keywords Air sampling Á Molds Á Traditional wine In these traditional wine cellars, all available sur- production faces—walls, barrels, bottles, etc.—are covered with a characteristic blackish-gray layer of Zasmidium (Cla- dosporium) cellare (noble mold, Fig. 1). This fungus has no moldy odor and allegedly contributes to the 1 Introduction creation of extraordinary climatic environment for faster wine aging and its bouquet. Local winemakers Microfungi often colonize subterranean spaces like suppose that the fungus filters out unwanted yeasts and wine cellars. Their spores are dominant components of bacteria and consumes airborne alcohol, aldehydes the indoor air. Surface and air samplings were carried and other organic substances evaporated during the out at different wine-producing areas in Austria aging of Tokaj wine (To´th 2006; Gasˇinec et al. 2012). (Clemenz et al. 2008; Haas et al. 2010), France The positive effects of Z. cellare have not been proven (Simeray et al. 2001), Italy and Switzerland (Picco and by scientific methods, and some superstitions aroused Rodolfi 2004), Japan (Goto et al. 1989), to determine related to its harmful effects on workers. Conse- fungal composition in wine cellars. In these studies, quently, Z. cellare has been removed recently in some several fungal species were detected, including cellars due to these hygiene considerations. Studies are pathogens. therefore required to evaluate whether there are any Some of them may occasionally have an influence positive or negative effects of fungal colonization and on the health of workers by inducing the development traditional versus new wine aging techniques on air of respiratory diseases (allergy, asthma), non-specific quality in wine cellars. To our knowledge, the fungi of respiratory symptoms and irritation (Douwes et al. wine cellars in Hungary have never been studied 2003; Fung and Hughson 2003; Flannigan et al. 2011; before, although fungi derived from grapes (e.g., Mendell et al. 2011); furthermore, their presence is Magyar and Bene 2006; Sipiczki 2016) and fermen- proposed as a cause of biodeterioration. Microfungi tation (e.g., Miklo´s et al. 1994; Sipiczki et al. 2001; habiting in wine cellars excrete volatile organic Naumov et al. 2002; Magyar and To´th 2011) are compounds that may also negatively affect the quality widely studied. of wine (Moularat et al. 2011). Tokaj is one of the The aim of this study was to collect data on the world’s great historical wine-producing regions, composition of the fungal biota of wine cellars in located in northeastern Hungary with strong viticul- Tokaj as well as their effect on air quality. ture traditions which have survived for over 1000 years (Nyizsalovszki and Fo´riań 2007; Boatto et al. 2011; UNESCO 2015). The vegetation period is 2 Materials and methods determined by the sunny, south-facing slopes and the proximity of the Tisza and Bodrog rivers. The wine The Tokaj wine region in Hungary consists of 27 quality depends largely on the specific climate of villages and 11,149 h of classified vineyards, of foothills being conducive to the proliferation of which an estimated 5500 h are currently planted. It Botrytis (noble rot) colonizing and subsequently lies in the close proximity of the northernmost desiccating the grapes that promote the production of climatic limit of wine production. The microclimate ‘aszu´’ (botrytized) wine. This process depends largely is cold and temperate (the average temperature is on meteorological conditions during September 9.9 °C). There is significant rainfall throughout the (Makra et al. 2009). The terroir consists of clay or year (averages 611 mm). Samples were collected in loess soil on volcanic subsoil (rhyolite) suitable for four communities in the Southern part of the Tokaj digging out the cellars without support system. The wine district with five traditional (subterranean) vast system of cellars with long and narrow corridors wine cellars (Tolcsva, Szegi, Tarcal 1–3) and one has been carved out of solid rock between 1400 and building with modern wine production technology 1600 AD. The cellars have long-term constant (Erdobe} ńye) in November 2012 (Table 1). Both 123 Aerobiologia

Fig. 1 a Landscape of Tokaj with the entrances of cellars, b Traditional wine cellar, c Colony of Zasmidium cellare on a wine bottle, d Colonies of Aspergillus spelunceus on the ﬂoor of a cellar. Photographs: a Bihari Z., b–d Magyar D outdoor and ambient conditions in each winery area 2.1 Examination of air quality were measured parallel with biological samplings (Tables 2 and 3, respectively). Air samples were collected in November 2012 at a height of 1.2 m above ground level with a radial diffusive sampler (RadielloÒ BTEX/VOC’s cartridge CS2 solvent desorption), activated with activated 123 Aerobiologia

Table 1 Characteristics of sampling locations Parameters Wine cellars AB CDEF

Location Tolcsva Erdobe} ´nye Szegi Tarcal Tarcal Tarcal Coordinates North 48°1602000 48°1505900 48°1202400 48°704200 48°704200 48°704400 East 21°2202900 21°2104200 21°2301900 21°2004700 21°2005400 21°2005400 Elevation (m) 135 299 146 104 181 110 Surroundingsa WW? FW? FM M M Age (years) 800 10 416 316 566 216 Reconstructed (years) 56 0 76 116 29 36 Depth (m) 15 0 30 13 5 11 Ventilationb 12 1131 Number of chimneys 20 0 10 6 0 40 Tunnels length (m) 700 0 5000 200 50 700 Tunnels width (m) 3 0 4 3 6 3 Mold removal (years) 55 1 46 10 6 200 Wine capacity (hl) 1600 800 20,000 200 140 60 Current wine load (%) 80 95 40 80 30 3 aThe surrounding locations W, F and M mean vineyard, forest and municipal areas, respectively. b1: ventilation passively through chimneys, 2: HVAC system, 3: no ventilation

Table 2 Environmental Parameters Wine cellars locations C.V.% parameters outdoors of the wine cellars sampled ABCDEF

Air pressure (hPa) 992.3 986.5 990.0 994.5 994.5 994.5 0.3 Temperature (°C) 15.0 15.5 12.4 13.1 13.1 13.1 9.1 Relative humidity (%) 59.3 58.1 67 62.7 62.7 62.7 5.0 Dew point (°C) 7.1 6.0 6.3 6.0 6.0 6.0 7.1 Evaporation temperature (°C) 10.7 9.9 9.1 9.3 9.3 9.3 6.3 The capital letters A–F Moisture content (g/kg) 6.4 6.0 6.0 5.9 5.9 5.9 6.0 mark wine cellars, see Absolute humidity (g/m3) 7.8 7.3 7.4 7.5 7.5 7.5 7.5 Table 1 charcoal (30–50 mesh, Fondazione Salvatore Mau- Reference air samples were collected outdoors, at geri, Tradate, Italia) to detect airborne ethanol and the entrance of wine cellars. methanol. Exposition time of tubes was 7 days. The samples were analyzed according to standards (MSZ 2.2.1 Surfaces EN 1076/1999, NIOSH 2000:1998, OSHA 100:1993, NIOSH 1405:200). Concentration levels were Surfaces were examined visually, and samples were expressed in lg/m3. collected from mold-covered walls and other objects with sterile cotton swabs and spread subsequently on 2.2 Examination of mycota MEA plates. The plates were incubated at 25 °C.

Two sampling techniques were used for the examination of fungal communities residing in wineries.

123 Aerobiologia

Table 3 Indoor conditions in wine cellars sampled Parameters Wine cellars C.V.% ABCDEF

Air pressure (hPa) 992.9 987.3 990.3 995.1 994.8 993.1 0.3 Temperature (°C) 17.7 16.5 14.8 14.6 13.8 13.7 10.5 Relative humidity (%) 74.1 52.4 72.9 85.4 89.5 84.8 17.7 dew point (°C) 13.1 6.5 9.9 12.3 12.7 11.2 22.5 evaporation temperature (°C) 14.8 10.9 12 13.3 13.5 12.3 10.6 moisture content (g/kg) 9.6 6.1 7.8 9.2 9.4 8.4 15.7 absolute humidity (g/m3) 11.4 7.6 9.5 11.1 11.3 10.2 14.4 Surface temperature (°C) min 12.3 18.3 10.8 14.1 13.6 11.6 19.9 Surface temperature (°C) max 14.3 14.3 11.3 17.0 14.1 12.4 14.0 Ethanol (lg/m3)a n.d. 136.8 206.6 16.8 105.2 1.6 91.4 Methanol (lg/m3)b n.d. n.d. n.d. n.d. n.d. n.d. – Average RBA (CFU/m3)c 2145 75 8190 2535 1825 1955 99.7 I/O ratio RBAd 0.65 0.05 9.05 1.40 1.01 1.08 153.4 The capital letters A–F mark wine cellars, see Table 1 a,b detection limits, a: 0.21 lg/m3, b: 0.17 lg/m3 c 3 average number of total colony-forming units (CFU) per m (Fcellar = 11.28, p = 0.042), samples collected onto Rose Bengal agar (RBA) (Freplication = 0.48, LSD0.05 = 350, p [ 0.1) d 3 indoor versus outdoor ratio of average number of total CFU per m (FI.O = 6.69, p \ 0.05) n.d.: not detected

2.2.2 Air samples factors according to Macher (1989) and expressed in CFU/m3 air. Air samples were taken 1.2 m above ground level, using a one-stage volumetric sieve sampler (SAS IAQ, 2.3 Isolation and maintenance of fungal strains International PBI S.p.A., Milan, Italy) at a flow of 100 l/min. Three types of culture media [Potato Strains representing m orphotypes were isolated from Dextrose Agar (PDA), malt extract agar (MEA, colonies formed on agar plates from either air or Samson 2010) and Askew agar with Rose Bengal surface samples and maintained on PDA slants at (RBA, Askew and Laing 1993)] supplemented with ambient temperature until their identification. chloramphenicol have been applied for sampling. The plates were incubated at 25 °C. Qualitative and 2.3.1 Molecular identification quantitative evaluation of CFUs (colony-forming units) were carried out following 3–8 days of cultur- For the molecular identification of the isolates, ing; 15 days for slow-growing species. CFUs were mycelia were grown in 250-ml flasks in 50 ml potato divided into different ‘morphotypes’ based on simi- dextrose broth (PDB, Scharlau Microbiology, 4 gL-1 larity of cultural characteristics, e.g., colony color, potato peptone, 20 gL-1 glucose in distilled water) by texture, and growth rates. Sporulating morphotypes shaking at 120 rpm at 20 °C for 10 days. The mycelial were identified at the genus level by their microscopic mat was ground to fine powder under liquid nitrogen and macroscopic morphology using standard myco- with a pestle in a mortar. Genomic DNA was isolated logical literature to validate the molecular identifica- with the GenEluteTM Plant Genomic DNA Miniprep tion (Raper and Fennell 1977; Von Arx 1981; Singh Kit (Sigma-Aldrich Ltd.) according to the manufac- 1991; Klich 2002; Samson 2010). The airborne fungal turer’s protocol. Then the internal transcribed spacer concentration was calculated using the correction (ITS) region of the rRNA gene cluster was amplified 123 Aerobiologia with primers ITS1 and ITS4 (White et al. 1990) and applying appropriate modules of Statistica 5 program sequenced. A Bioer Life ECO Thermal cycler was (StatSoft 5.0., Tusla, USA). Graphic plotting and used for PCR amplification with the following condi- regression analysis were used to disclose influence of tions: 2-min initial denaturation at 94 °C, followed by environmental factors (Tables 1, 2, 3); the acceptance 35 cycles of 1-min denaturation at 95 °C, 1-min level was set to 95% significance level. Details of primer annealing at 60 °C, 2-min elongation at 72 °C schemes followed were delineated previously (Mag- and a final extension at 72 °C for 15 min. The yar et al. 2011; Magyar and Oros 2012). The graphical amplified DNA products were purified using Gel/ presentation of the result of data analysis was edited PCR DNA Fragment Extraction Kit (Geneaid) accord- uniformly in MS Office PowerPoint 2003. ing to the manufacturer’s protocol. Sequencing was carried out by Microsynth Austria GmbH. With the sequences obtained, we carried out similarity searches 3 Results and discussion using the BLAST network service of the NCBI database (http://www.ncbi.nlm.nih.gov/blast.cgi). No extreme weather events happened in the season of After the database search, pairwise BLAST compar- sampling. The state of vegetation corresponded to isons were performed with each sequence and the usual conditions of the season. Surroundings of the sequences of the type strains of the species which gave cellars were well kept and typical for the area. Insides the most similar matches. When the ITS sequence outlooked and smelled as usual for the well-managed analysis gave ambiguous results, we also amplified wine cellars. Dark, spongious colonies of Zasmidium and sequenced the D1/D2 domains of the large-subunit (Cladosporium) cellare covered the walls of all 28S rRNA gene with the primers NL-1 and NL-4 subterranean cellars; however, it was not found in (O’Donnell 1993) under the following conditions: the modern building in Erdobe} ńye. The basidiomyce- 2-min initial denaturation at 94 °C, followed by 35 tous fungus Bjerkandera adusta (Polyporales) colo- cycles of 1-min denaturation at 94 °C, 1-min primer nized frequently some surfaces in subterranean cellars. annealing at 51 °C, 2.5-min elongation at 72 °C and a Also, red colonies of Aspergillus spelunceus covered final extension at 72 °C for 15 min. large areas on the floor of subterranean corridors of the cellar E. Visually distinguishable spots on surfaces 2.4 Data analysis were randomly found and sampled as well. Presence of 43 fungal species could be detected at the sampling Fisher’s test was applied to evaluate significance of sites (Table 4); however, none of them was repre- differences between variants (sampling, sampling sented in all samples. Moreover, some species points, wine cellars and composition of fungal occurred only in single samples. The difference consortia) at p = 0.05 level, and the interpretation between subterranean cellars and the modern building was made according to Sva´b(1979). The coefficient of is due to the diverse indoor conditions, especially variation [(100 9 LSD0.05)/Average)] was used for ventilation and surfaces types. Ventilation of the the characterization of individual parameters of modern building is mechanical, therefore indoor air ambient and outdoor environmental conditions. The quality of its open space could be controlled more data of observations were transformed into numerical effectively than that of the passively ventilated, values (1 and 0, presence or absence, respectively), irregularly carved tunnels of subterranean cellars and the resulted data matrix (in our case 18 sampling having numerous stagnant airspaces and dead ends. points and 43 fungal strains as varieties and observa- Surface types are also different: rough, porous and tions, respectively) was analyzed by multivariate often earthy and leaking (wet) walls and wood offer techniques to disclose patterns of distribution of better conditions for fungal growth in the subterranean species and similarities in diversity of fungal consortia cellars than concrete walls and containers from steel of in cellars. The dummy variables were used to avoid the modern building. heterogeneous errors of variation originated of alter- ations in sampling methods. Relationships between varieties were analyzed with nonlinear mapping (sampling points) and cluster analysis (fungal species) 123 Aerobiologia

Table 4 Fungal taxa No. Species Cellars isolated in wine cellars and surrounding outdoor air A B C D E F D–F In Out In Out In Out In In In Out

1 Absidia psychrophilia s– –– –– –––– 2 Acremonium brachypenium –– –– –– s––– 3 Alternaria alternata –– –– –– –a–– 4 Aphanocladium album a– –– –– –––– 5 Aspergillus niger –– –– –a a––– 6 Aspergillus spelunceus –– –– –– –as–– 7 Aspergillus varians –– –– as– –––– 8 Aspergillus versicolor –– –– –– a––– 9 Aspergillus spp. – – – – a – a – – a 10 Bjerkandera adusta –– –– –– ––s– 11 Botrytis cinerea –a –– –– aasa– 12 Candida membranifaciens s – – – as – – – – – 13 Cladosporium ossifragi –– a– –– –––– 14 Cladosporium spp. a a a a a a as a a a 15 Debaryomyces subglobosus a– –– –– –––– 16 Emericella nidulans a– –– –– ––a– 17 Epicoccum nigrum aa –a –a a–aa 18 Mortierella alpina s– –– –– –s–– 19 Mucor sp. –– –a a– ––a– 20 Penicillium adametzioides –– –– –– –a–– 21 Penicillium brevicompactum –– –– –– aa–– 22 Penicillium chrysogenum a– –– –– –––a 23 Penicillium citreonigrum –– –– as– a––– 24 Penicillium citrinum –– –– s– –––– 25 Penicillium commune –– –– a– –––– 26 Penicillium echinulatum –– –– –– ––as– 27 Penicillium expansum – – a – – – as a a – 28 Penicillium glandicola –– –– –– ––s– 29 Penicillium oxalicum –– –– –a –––a 30 Penicillium roqueforti –– –– –– s––– 31 Penicillium solitum –– –– –– ––a– 32 Penicillium spinulosum as – a – as – as as as – 33 Penicillium thomii as––––––a–– 34 Penicillium spp. a – a – a a a a a a 35 Phanerochaete chrysosporium –– –– –– s––– 36 Rasamsonia brevistipitata –– –– –– ––a– 37 Scopulariopsis sp. a– –– –– –––– 38 Talaromyces diversus –– –– –– –s–– 39 Talaromyces rugulosus –– –– –– ss–– The capital letters A–F mark wine cellars, see 40 Umbelopsis isabellina s– –– –– –––– Table 1 41 Zasmidium cellare s – – – s – as as s – a air sample, s surface 42 Yeasts a – – a a – as – – – sample, In indoor, Out 43 Non-sporulating strains as a a a s a as a a a outdoor sampling 123 Aerobiologia

3.1 Environmental conditions (old wooden barrels let more gaseous ethanol to leak than modern metal containers) and from the surfaces The outdoor conditions varied to a certain degree (due to the tradition of pouring wine onto the walls in (Table 2), however, the coefficient of variation (c.v.) Tokaj cellars) are the main sources of gaseous ethanol, did not surpassed 10%, indicating that there were no while minor amounts could be emitted by some fungal great differences in this respect. However, the indoor species, e.g., Penicillium expansum (as MVOC, see parameters varied at a higher degree (Table 3). Except Fiedler et al. 2001). The ethanol consumption of fungi from air pressure, the c.v. surpassed 10% having is also well known; Zasmidium cellare (Goodwin et al. maximum in the case of dew point. Correlations 2016) is seemingly the main utilizer in the sampled between outdoor and indoor parameters revealed cellars. Development of young colonies of Z. cellare dependence between several environmental factors. was often observed near ethanol sources, e.g., on corks The high correlation in air pressure (ri,o = 0.97, of wine bottles and near frequently used wine thieves. p \ 0.001) demonstrates the rapid equilibration that The effect of ethanol vapor on the growth of signalizes the perpetuous airflow between outer and Baudoinia compniacensis, a fungus colonizing build- inner airspaces. The absolute humidity also related ing walls near sources of ethanol emissions (e.g., positively but to a lesser extent (ri,o = 0.67, p \ 0.1), cognac aging warehouses, commercial bakeries and most probably due to the buffering capacity of insides. printing industry), was studied in detail (Scott et al. The latter can explain the absence of correlation in 2007; Ewaze et al. 2008). Amendment of growth temperature parameters. A study performed in tradi- medium with 5 ppm ethanol greatly increased the tional wine cellars in Spain showed that interior air success of isolation of this fungus. Optimal germina- temperature is fundamentally conditioned by the tion of dormant mycelia was seen at a low level of temperature of the ground and the outdoor air ethanol vapor (10 ppm). It was concluded that ethanol (Mazarroń and Canãs 2009). Ground temperature has may play a stimulatory role in the growth of colonies. a major influence. In the autumn and winter, the Zasmidium cellare and other frequent taxa may also influence of ground temperature decreases while that prefer high ethanol concentrations; therefore, further of the outside air increases, because the higher studies are needed to test whether the isolated fungi ventilation reduces the stability of the wine cellar. In have a more intensive growth on media supplemented the case of other parameters, the relationship also with ethanol. turned to be strictly linear (ri,o [ 0.84, p \ 0.05) when either cellar A or F outlying of regression was omitted 3.2.2 Density of fungal populations from the calculations. These two cellars differed from the others with high number of chimneys that under- The variation in CFU numbers on parallel plates in lines the dominant role of ventilation in the determi- samplers varied to a low extent (Frepl. = 0.41, nation of conditions in cellars. A Spanish study p [ 0.1); furthermore, the position of sampler in performed in the main wine-producing regions of the various locations caused seemingly negligible effect world showed that underground cellars have advan- on collection (Fexp = 1.64, p [ 0.1), supporting the tages for the aging and conservation of wine in all reliability of measurements. However, the number of regions studied, because of the thermal inertia of the countable colonies grown up on various selective ground and its ventilation behavior (Mazarroń et al., media placed in the sampler altered significantly 2012). (F = 6.69, p \ 0.05), which can be explained with differences in growth intensity of filamentous fungi on 3.2 Analytical determinations various media. Certain species (e.g., Alternaria, Botrytis, Epicoccum) rapidly overgrew the plate, 3.2.1 Airborne alcohol content suppressing the formation of visible colonies of other species. For this reason, the CFUs counted on RBA In all six wine cellars studied, airborne methanol was (Table 3) were taken as a measure of concentration of below the detection limits (Table 3). The aerial fungi in the air of the cellars as the fungal colonies ethanol content of cellars varied between 1.6 and developing on this medium showed moderate growth 206.8 lg/m3 (Table 3). Evaporation from the barrels rate (no overgrowing colonies). It has to be mentioned 123 Aerobiologia

Density of fungi the trafﬁc and the diverse ﬂora of parks and gardens in municipal areas are responsible for these differences. 1400 The outdoor samples were characterized by the

1200 presence of Cladosporium and Epicoccum spp., which were found in the air of the cellars as well. However, 1000 the frequency of propagules of fungal species in 3

/m 800 various sampling points varied substantially

600 (F = 24.43, p \ 0.001). For example, Botrytis CFU cinerea had high outdoor concentration in Tolcsva 400 only (140 CFU/m3 on RBA, but more common on 200 MEA and PDA: 640 and 870 CFU/m3, respectively), M while in Tarcal this fungus was not found in the 0 W Indoor Outdoor samples collected outdoors. Fig. 2 Influence of the environment around cellars on the Indoor samples were dominated by penicillia: a concentration of airborne fungi. M and W = munici- total of 15 species (including those referred to genera pal = downtown, wild = rural area Rasamsonia and Talaromyces) were present. The most frequent species was P. spinulosum detected in five that Rose Bengal, a member of the anthracene dye cellars; its highest concentration was above family, exhibiting noticeable antimicrobial activities 13,000 CFU/m3 in cellar C, while 47% of the peni- (Oros et al. 2003), can efficiently suppress both cillia were found in one cellar only; among them P. conidium germination and hyphal growth of fungal adametzioides, P. solitum and P. thomii had high species from the orders Sphaeriales and Helotiales levels (710, 2120 and 760 CFU/m3 in cellars A, E and (Oros and Cserha´ti 2009); consequently, the densities F, respectively). Penicillium expansum, being absent of Cladosporium and Botrytis in the sampled area in outdoor samples collected in Tarcal, was common might be underestimated. Furthermore, spreading the in the indoor samples collected in the cellars of this swab samples directly on MEA plates may underes- village and had also low level in cellar C. The genus timate fungal diversity, as the dominant species may Penicillium was richest in species in Tokaj, but also in suppress slow-growing taxa. The sampling method certain wineries in Austria (Haas et al. 2010), France allowed an insight into the fungal composition of the (Simeray et al. 2001), Italy and Switzerland (Picco and cellars, but several taxa remained undiscovered by the Rodolfi 2004) and Japan (Goto et al. 1989). Penicil- culturing method, as it is suggested by direct micro- lium expansum, Botrytis cinerea and some other scopical observations of the surfaces (i.e., tape lift fungal species could produce geosmin (trans-1,10- samples or plating serial dilutions onto different dimethyl-trans-9-decalol, La Guerche et al. 2005), culture media including media supplemented with while some Penicillium (and Trichoderma) species ethanol). Therefore, future studies should be focused may contribute to the formation of 2,4,6-trichloroani- on the development of a more efficient technique sole (TCA, Prak et al. 2007). These fungal metabolites based on monospore DNA analysis. could cause bad flavor of wines. Geosmin is an earthy- Significant differences were disclosed in CFU musty compound. TCA contamination is recognized numbers in the air of six cellars (Fcellars = 11.28; by cork taint, the most frequent wine fault (’wet dog or p \ 0.01), and to a lesser degree between CFU wet cardboard smell’), which makes wine mostly numbers of the samples (Foutdoors = 11.28; unpalatable (Tanner and Zanier 1981). TCA was p \ 0.01) collected in their outdoors. Significant detected by Haas and coworkers (2010) in the air of differences were found also between the concentra- Austrian wine cellars with visible moldy patches. tions of airborne propagules measured outdoors and Further studies are therefore needed to see whether inside cellars (Fo.i = 5.86; p \ 0.05). The concentra- these metabolites are present in Tokaj wine cellars. tion of airborne propagules measured either outside or Two members of the genus Aspergillus were also inside the cellars was significantly higher in municipal common inside the Tokaj cellars (A. nidulans in cellar areas than in rural ones (Fig. 2). One can assume that A: 110 and F: 90 CFU/m3; A. spelunceus in cellar E: 490 CFU/m3). Moreover, Aphanocladium album was 123 Aerobiologia an abundant fungus but only in cellar A (980 CFU/ Fig. 4 Cluster-diagram of the interrelationships among fungal c m3). Among yeasts, Candida membranifaciens (in species presented in Tokaj wine cellars. Grouping of species 3 based on their presence (?) or absence (-) in sampling points cellar C: [ 13,000 CFU/m ) and Debaryomyces han- (Table 4). Unweighted pair group average method based on 3 senii (in cellar A: 760 CFU/m ) were detected in high Pearson’s correlation coefficients was applied to construct the numbers. clusterogram. Codes of cellars (A-F) are the same as in Table 1. The samplings revealed high diversity of fungi A and B: superclusters; a1-3, b1-6: subclusters. The species marked with symbol filled circle produce hydrophilic spores. attached to the surfaces in traditional cellars. Most of The black, gray and white prisms on the histogram refer to the the isolated strains have been identified at the species number of species present in surface, indoor and outdoor air level (Table 4). Besides the aforementioned Z. cellare, samples, respectively A. spelunceus and B. adjusta, 18 other species occurred in surfaces, among them 11 were found in air samples as well. Penicillium spinulosum was identified in airspaces of cellars located in municipal present in all traditional cellars, although its incon- areas, where common species dominated over rare spicuous colonies could not be perceived with naked ones. Although the number of propagules was the eyes before the evaluation of the samples. Although Z. highest in outdoor air of municipal areas (Fig. 2), the cellare was present in all subterranean cellars, it was species diversity was below the one detected at other detected in the air of only two of them (in cellar D and sampling points; moreover, the number of sporadic 3 E: 30 and 10 CFU/m , respectively). Simeray et al. species was also the lowest in this case (Fig. 3). (2001) had similar results in French cellars, where Zasmidium mycelia commonly cover the walls but the 3.3 Diversity of fungal consortia fungus is rarely isolated from the air samples, possibly due to poor sporulation. The MANOVA of population density data revealed As demonstrated in Fig. 3, the outside environment significant differences (p \ 0.01) between either out- did not significantly influence the number of species door or indoor sampling points of various cellars habiting on surfaces of cellars located either in rural or (Table 3). To disclose patterns of species distribution municipal areas; the ratio of frequent and sporadic and similarities in the diversity of fungal consortia in species was similar in both locations with higher cellars, the data of Table 4 were transformed into number of the latter group. Contrarily, great differences numeric values on the basis of presence/absence of manifested in the composition of airborne species in species to correct for heterogeneous variances origi- cellars: the consortia indoors were more diverse than nating from differences in samplings. The resulting outdoors, and the highest number of species was data matrix was subjected to nonlinear mapping (Sammon 1969) and cluster analysis (unweighted pair Diversity of fungi group average method based on Pearson’s correlation

15 coefficients). The species formed two large clusters (Fig. 4), both of them enclosing several subclusters 10 (p = 0.01 - 0.05). The first supercluster (A), dominated by airborne species, comprises mainly fungi producing dry (hydrophobic) spores, and only one of 5 the three subclusters included some species having Number of species readily wettable (hydrophilic) spores. Spores grouped M according to their wettability (see Mason 1937; 0 W Surface Indoor Outdoor Magyar et al. 2016). Airspaces Most species isolated from surfaces belonged to the other supercluster (B), where three out of six subclus- Fig. 3 Influence of the environment around cellars on the diversity of airborne fungi. M and W = municipal = down- ters included only species producing hydrophobic town, wild = rural area. The lower (dark gray) and upper (light spores. The majority of the hydrophilic-spored species gray) parts of the columns are proportional to the number of agglomerated into other three subclusters was isolated frequent and rare species trapped 123 Aerobiologia

from surface samples, and most of them were found hydrophylic spores and these spores could hardly be only in one cellar. Hydrophilic species have aerosolized (they liberate if the surface dries out).

123 Aerobiologia

Consequently, their airborne dispersal could be lim- Among the environmental conditions, elevation, age, ited in the moist environment of subterranean cellars. reconstruction time of cellars, indoor ethanol concen- Fungal diversity in various sampled locations, i.e., tration and number of chimneys influenced the diversity on surfaces and in either indoor or outdoor air was of mycota of the examined wine cellars to a certain significantly different (F = 12.19, p \ 0.001). degree (p \ 0.05). Except for elevation, these param- Assuming that propagules collected demonstrate the eters had no effect on the fungal diversity of the outdoor diversity of resident fungal consortia in situ, the air (p [ 0.1). Elevation had significant negative effect position of sampling points plotted as variables of on the diversity of fungi, both outdoors (R2 = 0.801) nonlinear map (Fig. 5) relates to the similarity of and indoors (air R2 = 0.787; surface R2 = 0.779). The fungal populations in the proper locations, i.e., dis- level of relative humidity also influenced the diversity tances between points are proportional to their simi- (R2 = 0.627), which was significantly lower in the larity. The sampling locations formed well-separated cellars at higher elevation; most probably the lower clusters, among them the outdoor air samples being the amount of water offered for fungal growth negatively most compact one. The group of indoor airspaces affected the fungal diversity in these cellars. Indoor formed a less compact cluster with strictly separated humidity is high in traditional or partially subterranean cellar C, which is the most spacious among the cellars spaces, like wine cellars and dwellings, where it could examined. The surface samples of various cellars pose problems if the presence of molds is not controlled showed low similarity to the respective airspaces. (Canãs and Ocanã 2005; Malaktou et al. 2016). The species diversity both on the cellar surfaces and 3.4 Factors influencing the fungal consortia in the indoor airspaces positively correlated with the in wine cellars age of cellars (R2 = 0.529 and 0.870, respectively). The age after reconstruction—similarly to the age of The data on density and fungal diversities were plotted the cellars—also positively correlated with the against environmental variables (Tables 1, 2, 3), and increase in fungal diversity both on surfaces regression analysis was applied to evaluate the (R2 = 0.659) and in airspaces (R2 = 0.742), indicat- strength of relationships. ing the rapid regeneration of characteristic mycobiota

Fig. 5 Nonlinear map of sampling points. The size of circles is proportional to the number of fungal species collected in the sampling sites. The samples collected in either outdoor or indoor airspaces are ﬁlled with white and gray while those of surfaces with black. The capital letters A–F mark wine cellars (see Table 1)

123 Aerobiologia in cellars. Regarding the diversity of the fungi growing Subterranean cellars ventilate through a door and on the cellar surface, elevation had about tenfold chimneys; however, the number of chimneys varied stronger effect than age, but was comparable with markedly by the cellars studied. The number of reconstruction time. The modern wine store was chimneys affected the diversity of fungi positively in excluded from this analysis, since its unique features the indoor air (R2 = 0.68) but not outdoors. Especially were not comparable with the subterranean cellars. the number of sporadic species increased in the air of Since Zasmidium coverhasaculturalrolein the cellars. Interestingly, the effect was negative on the traditional wine making and also attracts tourists, the diversity of surface mycota (R2 = 0.71). It is thought importance of such studies should be emphasized. The that air currents may transport spores from the outdoor reconstruction of the cellars included the removal of environment through doors, adding new components Zasmidium cover from the walls. This process decreased to the mycota of the indoor air, but these fungi were both the diversity (p \ 0.05) and the density (p \ 0.001) not able to establish colonies on the walls due to their of airborne and surface fungi in the indoor airspace. low competitive ability against the resident mycota. Nevertheless, the resident fungal consortia recovered More chimneys may reduce the relative humidity and seem to be stable; according to the trend calculated indoors by a better ventilation and make conditions of the regression equation, constant members present in unfavorable for fungal growth on the surfaces. thearea(Cladosporium, Epicoccum, Penicillium, Zas- midium and yeasts) rapidly start to colonize the subterranean surfaces. The old fungal cover can contribute to 4 Conclusions the species richness of the surfaces in the cellars sustaining favorable condition for microbiota. Zasmid- There were marked differences in the airborne con- ium has a large surface area of mycelia but low spore centration and composition of fungi in the examined production. We suppose that this Zasmidium cover acts wine cellars. Apparently, two factors have major as an air filter, trapping spores of other fungi. A fraction effects on the fungal population: concentration of of the trapped fungal species may survive in the airborne ethanol (as nutrient and as an agent promoting Zasmidium colonies as competitors or mycoparasites the competitiveness of certain species) and ventilation (unpublished observations of the authors). Some of these (number of chimneys, temperature and air pressure). fungi can release large numbers of spores, which become The number of spore-forming species generated high airborne and may trigger allergy and asthma (Day and diversity of indoor fungi and differences between the Ellis 2001). However, no medical cases were docu- cellars’ fungal compositions; however, their dominant mented in theTokaj region and no experience gave either species proved to be the same. rise to belief about the health hazard caused by fungal cover on cellar surfaces. Ethanol can promote the growth of not only 5 Acknowledgements Zasmidium but also of other fungal species that have abundant spore production. The concentration of fungi La´szlo´ Kredics was supported by grant GINOP-2.3.2- in the air was positively correlated (p \ 0.005) with 15-2016-00012 (Szećhenyi 2020 Programme) and the the airborne concentration of ethanol; nevertheless, Jańos Bolyai Research Scholarship (Hungarian Acad- the latter poorly influenced (p = 0.347) the diversity emy of Sciences). The research was carried out partly of mycota of cellars. Our results did not confirm the with the support of OTKA K13255. hypothesis that the higher ethanol level decreases airborne spore concentration by intensifying the growth of the non-sporulating but spore-trapping Zasmidium on the walls. However, further studies References are needed to find out whether the removal of the Zasmidium cover is beneficial or not, moreover, how a Askew, D. J., & Laing, M. D. (1993). An adapted selective medium for the quantitative isolation of Trichoderma homogeneous Zasmidium cover can be created with- species. Plant Pathology, 42(5), 686–690. out sporulating ‘partner species’. Boatto, V., Defrancesco, E., & Trestini, S. (2011). The price premium for wine quality signals: does retailers’ 123 Aerobiologia

information provision matter? British Food Journal, Magyar, I., & Bene, Z. (2006). Morphological and taxonomic 113(4–5), 669–679. study on mycobiota of noble rotted grapes in the Tokaj Canãs, I., & Ocanã, S. M. (2005). Study of the thermal beha- wine district. Acta Alimentaria, 35(2), 237–246. viour of traditional wine cellars: the case of the area of Magyar, D., Esze´ki, E. R., Oros, G., Szećsi, A´ ., Kredics, L., ‘‘Tierras Sorianas del Cid’’ (Spain). Renewable Energy, Hatvani, L., et al. (2011). The air spora of an orchid 30(1), 43–55. greenhouse. Aerobiologia, 27(2), 121–134. Clemenz, A., Sterflinger, K., Kneifel, W., & Mandl, K. (2008). Magyar, D., & Oros, G. (2012). Application of the Principal Airborne fungal microflora of selected types of wine-cel- Component Analysis to disclose factors influencing on the lars in Austria. Mitteilungen Klosterneuburg, 58, 17–22. composition of fungal consortia deteriorating remained Day, J. H., & Ellis, A. K. (2001). Allergenic microorganisms fruit stalks on sour cherry trees. In P. Sanguansat (Ed.), and hypersensitivity. In B. Flannigan, R. A. Samson, & J. Principal Component Analysis - Multidisciplinary Appli- D. Miller (Eds.), Microorganisms in Home and Indoor cations (pp. 89–110). Rijeka: InTech. Work Environments. Diversity, Health Impacts, Investi- Magyar, I., & To´th, T. (2011). Comparative evaluation of some gation and Control. Boca Raton, London, New York: CRC oenological properties in wine strains of Candida stellata, Press, Taylor & Francis Group. Candida zemplinina, Saccharomyces uvarum and Sac- Douwes, J., Thorne, P., Pearce, N., & Heederik, D. (2003). charomyces cerevisiae. Food Microbiology, 28(1), Bioaerosol health effects and exposure assessment: pro- 94–100. gress and prospects. Annals of Occupational Hygiene, 47, Magyar, D., Vass, M., & Li, D. W. (2016). Dispersal strategies 187–200. of microfungi. In Biology of Microfungi (pp. 315–371). Ewaze, J. O., Summerbell, R. C., & Scott, J. A. (2008). Ethanol Springer International Publishing. physiology in the warehouse-staining fungus. Baudoinia Makra, L., Vitańyi, B., Ga´l, A., Mika, J., Matyasovszky, I., & compniacensis. Mycological Research, 112(11), Hirsch, T. (2009). Wine quantity and quality variations in 1373–1380. relation to climatic factors in the Tokaj (Hungary) wine- Fiedler, K., Schu¨tz, E., & Geh, S. (2001). Detection of microbial growing region. American Journal of Enology and Viti- volatile organic compounds (MVOCs) produced by culture, 60(3), 312–321. moulds on various materials. International Journal of Malaktou, E., Philokyprou, M., Michael, A., & Savvides, A. Hygiene and Environmental Health, 204(2–3), 111–121. (2016). Thermal assessment of traditional, partially sub- Flannigan, B., Samson, R. A., & Miller, J. D. (2011). Mi- terranean dwellings in coastal and mountainous regions in croorganisms in home and indoor work environments. the Mediterranean climate. The case of Cyprus. Journal of Diversity, health impacts, investigation and control (2nd Sustainable Architecture and Civil Engineering, 16(3), ed.). Boca Raton, London, New York: CRC Press, Taylor 82–96. & Francis Group. Mason, E. W. (1937). Annotated account of fungi received at the Fung, F., & Hughson, W. G. (2003). Health effects of indoor Imperial Mycological Institute. (List II, Fasc.3-General fungal bioaerosol exposure. Applied Occupational and part) Kew Surrey: IMI. Environmental Hygiene, 18(7), 535–544. Mazarroń, F. R., & Canãs, I. (2009). Seasonal analysis of the Gasˇinec, J., Ra´kay, Sˇ., Ra´kay, Sˇ., & Sˇimcˇak, M. (2012). thermal behaviour of traditional underground wine cellars Geodetic survey of tuff wine cellars in Vel’ka Trńˇa. Acta in Spain. Renewable Energy, 34(11), 2484–2492. Montanistica Slovaca, 17(3), 189–194. Mazarroń, F. R., Cid-Falceto, J., & Canãs, I. (2012). An Goodwin, S. B., McCorison, C. B., Cavaletto, J. R., Culley, D. assessment of using ground thermal inertia as passive E., LaButti, K., Baker, S. E., et al. (2016). The mitochon- thermal technique in the wine industry around the world. drial genome of the ethanol-metabolizing, wine cellar mold Applied Thermal Engineering, 33–34, 54–61. Zasmidium cellare is the smallest for a filamentous asco- Mendell, M. J., Mirer, A. G., Cheung, K., Tong, M., & Douwes, mycete. Fungal Biology, 120(8), 961–974. J. (2011). Respiratory and allergic health effects of Goto, S., Takayama, K., & Shinohara, T. (1989). Occurence of dampness, mold, and dampness-related agents: a review of molds in wine storage cellars. Journal of Fermentation and the epidemiologic evidence. Environmental Health Per- Bioengineering, 68(4), 230–232. spectives, 119(6), 748–756. Haas, D., Galler, H., Habib, J., Melkes, A., Schlacher, R., Miklo´s, I., Sipiczki, M., & Benko, Z. (1994). Osmotolerant Buzina, W., et al. (2010). Concentrations of viable airborne yeasts isolated from Tokaj wines. Journal of Basic fungal spores and trichloroanisole in wine cellars. Inter- Microbiology, 34(6), 379–385. national Journal of Food Microbiology, 144(1), 126–132. Moularat, S., Hulin, M., Robine, E., Annesi-Maesano, I., & Klich, M. A. (2002). Identification of common Aspergillus Caillaud, D. (2011). Airborne fungal volatile organic species. Utrecht: Centraalbureau voor Schimmelcultures. compounds in rural and urban dwellings: detection of La Guerche, S., Chamont, S., Blancard, D., Dubourdieu, D., & mould contamination in 94 homes determined by visual Darriet, P. (2005). Origin of (-)-geosmin on grapes: on the inspection and airborne fungal volatile organic compounds complementary action of two fungi, Botrytis cinerea and method. Science of the Total Environment, 409(11), Penicillium expansum. Antonie van Leeuwenhoek, 88(2), 2005–2009. 131–139. Naumov, G. I., Naumova, E. S., Antunovics, Z., & Sipiczki, M. Macher, J. M. (1989). Positive-hole correction of multiple-jet (2002). Saccharomyces bayanus var. uvarum in Tokaj impactors for collecting viable microorganisms. The wine-making of Slovakia and Hungary. Applied Microbi- American Industrial Hygiene Association Journal, 50(11), ology and Biotechnology, 59(6), 727–730. 561–568. 123 Aerobiologia

Nyizsalovszki, R., & Fo´riań, T. (2007). Human impact on the Simeray, J., Mandin, D., Mercier, M., & Chaumont, J.-P. (2001). landscape in the Tokaj foothill region Hungary. Geografia Survey of viable airborne fungal propagules in French wine Fisica e Dinamica Quaternaria, 30(2), 219–224. cellars. Aerobiologia, 17, 19–24. O’Donnell, K. (1993). Fusarium and its near relatives. In D. Singh, K. (1991). An illustrated manual on identification of R. Reynolds & J. W. Taylor (Eds.), The Fungal Holo- some seed-borne Aspergilli, Fusaria, Penicillia and their morph: Mitotic, Meiotic and Pleomorphic Speciation in mycotoxins. Hellerup: Danish Government Institute of Fungal Systematic (pp. 225–233). Wallingford: CAB Seed Pathology for Developing Countries. International. Sipiczki, M. (2016). Overwintering of vineyard yeasts: survival Oros, G., & Cserha´ti, T. (2009). Use of principal component of interacting yeast communities in grapes mummified on analysis and spectral mapping technique for the evaluation vines. Frontiers in Microbiology, 7, 212. of the antifungal activity of anthracene-based synthetic Sipiczki, M., Romano, P., Lipani, G., Miklos, I., & Antunovics, dyes. SAR & QSAR in Environmental Sciences, 20(3), Z. (2001). Analysis of yeasts derived from natural fer- 379–391. mentation in a Tokaj winery. Antonie van Leeuwenhoek, Oros, G., Cserha´ti, T., & Forgaćs, E. (2003). Separation of the 79(1), 97–105. strength and selectivity of the microbiological effect of Sva´b, J. (1979). Multivariate methods in biometrics. Budapest: synthetic dyes by spectral mapping technique. Chemo- Mezogazdasa} ´gi Kiado´. (in Hungarian). sphere, 52(1), 185–193. Tanner, H., & Zanier, C. (1981). Zur analytischen Differen- Picco, A. M., & Rodolfi, M. (2004). Assesment of indoor fungi zierung von Muffton und Korkgeschmack in Weinen. in selected wineries of Oltrepo Pavese (Northern Italy) and Schweizer Zeitschrift fu¨r Obst und Weinbau, 117, 752–757. Sottoceneri (Switzerland). American Journal of Enology To´th, D. (2006). Diversity of microflora in Tokaj cellars, Tokaj and Viticulture, 55(4), 355–362. winegrowing and winemaking in Slovakia: almanac Prak, S., Gunata, Z., Guiraud, J.-P., & Schorr-Gallindo, S. account from specialized seminar activity from 19. (2007). Fungal strains isolated from cork stoppers and the November 2002 in Vinicky, Agrogenofond, Nitra. formation of 2,4,6-trichloroanisole involved in the cork UNESCO description: Tokaj Wine Region Historic Cultural taint of wine. Food Microbiology, 24, 271–280. Landscape. http://whc.unesco.org/ Accessed 03 November Raper, K. B., & Fennell, D. I. (1977). The Genus Aspergillus. RE 2015. Krieger Publishing Company. New York: Huntington. Von Arx, J. A. (1981). The genera of fungi sporulating in pure Sammon, J. W. (1969). A nonlinear mapping for data structure culture (No. Ed. 3). Vaduz: J. Cramer. analysis. IEEE Transactions on Computers, 100(5), White, T. J., Bruns, T., Lee, S., & Taylor, J. (1990). Amplifi- 401–409. cation and direct sequencing of fungal ribosomal RNA Samson, R. A. (2010). Food and Indoor Fungi. CBS Laboratory genes for phylogenetics. In M. H. Innes, D. H. Gelfand, J. Manual Series. Utrecht: CBS-KNAW Fungal Biodiversity J. Sninsky, & T. J. White (Eds.), PCR protocols (pp. Centre. 315–322). San Diego: Academic Press. Scott, J. A., Untereiner, W. A., Ewaze, J. O., Wong, B., & Doyle, D. (2007). Baudoinia, a new genus to accommodate Torula compniacensis. Mycologia, 99(4), 592–601.

123